One of our long-term goals is to understand how base excision repair (BER) maintains genetic integrity and functions in epigenetic regulation. This project aims to reveal how DNA glycosylases, which initiate BER, recognize and remove deaminated or oxidized forms of 5-methylcytosine (mC) from DNA, and how their activity is regulated by post-translational modification. As the most abundant modified base in DNA, mC serves as an epigenetic mark for gene silencing in eukaryotes and functions in the restriction modification systems of archaea and bacteria. However, cytosine methylation also poses a serious threat to genetic and epigenetic integrity. Deamination of mC to T generates G/T mispairs, and, upon replication, C to T transitions. Through this process mC deamination causes a large fraction of point mutations in cancer and genetic disease, which highlights the need to understand how glycosylases recognize and process G/T mispairs. Three types of glycosylases initiate repair of G/T mispairs: TDG (thymine DNA glycosylase), MBD4 (methyl binding domain IV), and MIG (mismatch glycosylase). While most glycosylases remove bases that are foreign to DNA (e.g., uracil), these mismatch enzymes remove thymine from rare G/T mispairs but not from the huge background of A:T pairs in DNA. Because aberrant glycosylase action on undamaged DNA is mutagenic and cytotoxic, the specificity of these enzymes is critical, but it remains poorly understood. The current paradigm holds that mismatch specificity derives from enzyme contacts with the mismatched guanine, but this remains unsubstantiated. Using a synergistic combination of experimental and computational methods, we will test this model and investigate three other potential specificity factors, defining the mechanism of G/T mismatch specificity for MBD4, MIG, and TDG. Recent studies show that TET enzymes oxidize mC, to 5-hydroxymethyl- C (hmC), 5-formyl-C (fC), and 5-carboxyl-C (caC), and that TDG excision of fC (or caC) and follow-on BER completes a TET-TDG-BER pathway for DNA demethylation. This key function in epigenetic regulation likely explains the essentiality of TDG for embryogenesis. However, the molecular basis of this recently discovered activity is poorly defined. New high-resolution crystal structures will reveal how TDG recognizes fC and caC in DNA. We will also investigate the molecular basis of new findings that TDG exhibits a vast difference in base- pairing preferences for excision of fC and caC relative to T (and other uracils), and investigate the capacity of TDG and BER to process proximal fC or caC bases in opposite strands without generating proximal AP sites or double-strand breaks. TDG is subject to SUMO modification and it has a SUMO-interacting motif (SIM) that binds non-covalently to SUMO domains. Mechanisms that regulate levels of SUMO~TDG in cells are poorly understood. The proposed studies will reveal isopeptidases that regulate TDG, and define their specificity and efficiency. A novel in vitro SUMO modification-deconjugation system will be used to directly test the current paradigm that sumoylation of TDG regulates product release and enhances enzymatic turnover.